Using EHR-QC to preprocess electronic medical records for data modelling

This tutorial is presented at the;

  • AI+Care conference 2023

  • Health Informatics and Knowledge Management (HIKM’24)

Overview

The integration of electronic health records (EHRs) has opened new avenues for leveraging historical data in predicting clinical outcomes and enhancing patient care. Nonetheless, the existence of non-standardized data formats and anomalies poses significant hurdles in utilising EHRs for digital health research. Additionally, to develop robust and reproducible predictive models, one needs to use data from multiple healthcare sites to account for population-wide variations in their modelling approaches. However, institution-specific data formats and inherent heterogeneity of EHR data hinder seamless data harmonisation. To tackle these issues head-on, we introduce EHR-QC, a comprehensive tool comprising two core modules: the Data Standardization Module and the Pre-processing Module.

Instructors

  1. Associate Professor Sonika Tyagi

Sonika is a group leader and an Associate Professor of Digital Health at the School of Computational Technologies, RMIT University. Sonika is also a co-investigator in the SuperbugAI flagship project at the Central Clinical School Monash University.

  1. Mr. Yashpal Ramakrishnaiah

Yashpal is a PhD student at the Central Clinical School under the supervision of Sonika Tyagi. Yashpal is the primary developer of the EHR-QC toolkit.

Learning Objectives

  1. Identifying challenges in applying EHR in a research setting

  2. Exploring medical records using EHR-QC modules

  3. Transforming raw medical records to standard data models or schemas

  4. Performing medical concept mapping using a standard vocabulary of choice

  5. Exploring various options to clean and pre-process data to make it machine-learning-ready

Prerequisite

Our target audience is biomedical and health data scientists and clinical practitioners. For conceptual understanding of the process, no prior knowledge of the subject is required. To participate in the hands-on tutorial attendees are required to have a prior knowledge of Python coding.

Tool dependencies

  1. Python 3.8 or above

  2. Postgres DB (required only for data standardisation, but not for preprocessing)

  3. Docker (optional)

  4. Access to any EHR data in csv format

Timetables

AI+Care conference 2023

Activity

Timings

Lecture

23 Nov 2023 11:00 AM to 11:15 AM

Live Demo

23 Nov 2023 11:15 AM to 11:40 AM

Q&A

23 Nov 2023 11:40 AM to 11:45 AM
Health Informatics and Knowledge Management (HIKM’24)

Activity

Timings

Introduction

01 Feb 2024 03:30 PM to 03:45 PM

Live Demo

01 Feb 2024 03:45 PM to 04:20 PM

Q&A

01 Feb 2023 04:20 AM to 04:30 PM

If you use this workflow please cite the following journal article

Contents